Thrombolysis in Myocardial Infarction Risk Score for Secondary Prevention of Recurrent Cardiovascular Events in a Real-World Cohort of Post-Acute Myocardial Infarction Patients

Duo Huang; Yang-Yang Cheng; Yiu-Tung Anthony Wong; See-Yue Arthur Yung; Chor-Cheung Frankie Tam; Ki-Wan Kelvin Chan; Cheung-Chi Simon Lam; Kai-Hang Yiu; Jo Jo Hai; Chu-Pak Lau; Esther W Chan; Chern-En Chiang; Ka-Lam Wong; Tommy Cheung; Bernard Man-Yung Cheung; Ying-Qing Feng; Ning Tan; Ji-Yan Chen; Wen-Sheng Yue; Hou-Xiang Hu; Ling Chen; Hung-Fat Tse; Pak-Hei Chan; Chung-Wah Siu

doi:10.1253/circj.CJ-18-0308

Abstract

Background: Patients who survive myocardial infarction (MI) are at risk of recurrent cardiovascular (CV) events. This study stratified post-MI patients for risk of recurrent CV events using the Thrombolysis in Myocardial Infarction (TIMI) Risk Score for Secondary Prevention (TRS 2°P).

Methods and Results: This was an observational study that applied TRS 2°P to a consecutive cohort of post-MI patients. The primary outcome was a composite endpoint of CV death, non-fatal MI, and non-fatal ischemic stroke. A total of 1,688 post-MI patients (70.3±13.6 years; male, 63.1%) were enrolled. After a mean follow-up of 41.5±34.4 months, 405 patients (24.0%) had developed a primary outcome (9.3%/year) consisting of 278 CV deaths, 134 non-fatal MI, and 33 non-fatal strokes. TRS 2°P was strongly associated with the primary outcome. The annual incidence of primary composite endpoint for patients with TRS 2°P 0 was 1.0%, and increased progressively to 39.9% for those with TRS 2°P ≥6 (HR, 27.6; 95% CI: 9.87–77.39, P<0.001). The diagnostic sensitivity of TRS 2°P for the primary composite endpoint was 76.3% (95% CI: 72.1–80.5%). Similar associations were also observed between TRS 2°P and CV death and non-fatal MI, but not non-fatal ischemic stroke.

Conclusions: TRS 2°P reliably stratified post-MI patients for risk of future CV events.

Atherosclerotic cardiovascular (CV) disease is currently one of the most common causes of morbidity and mortality worldwide.¹ Patients who survive a major CV event such as myocardial infarction (MI) are considered at highest risk of a future major CV event and should therefore be managed very aggressively.¹^–³ This includes prolonged dual antiplatelet therapy,⁴ potent statin therapy,³ and non-statin therapy, ezetimibe,⁵^,⁶ as well as certain novel therapeutic agents such as proprotein convertase subtilisin/kexin type 9 inhibitors,⁷ and the glucagon-like peptide 1 analogue, liraglutide, in patients with type 2 diabetes mellitus.⁸ Although these therapies have been shown to effectively reduce recurrent CV events, they may be accompanied by adverse effects such as bleeding and may also be prohibitively expensive.⁹^–¹¹ As such, accurate assessment of the absolute risk of future major CV events to identify very high-risk patients will allow efficient and appropriate use of these agents. The risk of future CV events such as recurrent MI, stroke and CV death in patients who survive MI nonetheless varies widely. In primary prevention, several CV risk-stratification schemes are available to guide long-term clinical decisions, such as Framingham risk score,¹² SCORE,¹³ and QRISK.¹⁴ In contrast, existing risk stratification models for secondary prevention, for example in patients with recent acute coronary syndrome (ACS), focus largely on short-term prognostication and do not assist long-term therapeutic decision-making.

Editorial p 713

Recently, the Thrombolysis in Myocardial Infarction (TIMI) group has developed a simple 9-point risk stratification tool to predict recurrent CV events using data from the Thrombin Receptor Antagonist in Secondary Prevention of Atherothrombotic Ischemic Events, TIMI 50 trial that involved stable patients with previous MI.¹⁵^,¹⁶ This TIMI Risk Score for Secondary Prevention (TRS 2°P) incorporates 9 readily available clinical characteristics: congestive heart failure (CHF), hypertension, diabetes mellitus, age ≥75 years, prior stroke, prior coronary artery bypass graft (CABG), peripheral artery disease (PAD), estimated glomerular filtration rate (eGFR) <60, and smoking and is able to identify an approximately 5-fold gradient in the risk of recurrent major CV events across low-, intermediate-, and high-risk categories.¹⁶ Using data from the Improved Reduction of Outcomes: Vytorin Efficacy International Trial (IMPROVE-IT),¹⁷ the score has been further validated for prediction of recurrent CV events following ACS in patients with established coronary artery disease (CAD). To date, TRS 2°P has been primarily validated in patient cohorts from randomized controlled trials (RCT), and its relevance to real-world clinical practice, particularly in different ethnicities, is uncertain. The aim of this study was therefore to describe the risk of major CV events in a large contemporary real world cohort of post-MI patients with detailed long-term follow-up, and to assess the prognostic performance of TRS 2°P in this population.

Methods

Study Design and Patients

This was an observational study using a hospital-based acute MI registry.¹⁸ The study protocol was approved by the local institutional review board. Informed consent was not obtained from patients given the registry nature of the study; nonetheless all patient records/information were anonymized prior to analysis. Patients diagnosed with acute MI based on the standard 9th International Classification of Disease (ICD-9) codes of 410.01, 410.11, 410.21, 410.31, 410.41, 410.51, 410.61, 410.71, 410.81, 410.91 and/or 411.1 at Queen Mary Hospital, Hong Kong, from February 2004 to January 2011, were identified via the computerized database of the clinical management system. Patients who died during hospitalization for the index MI were excluded from the final analysis. In Hong Kong, all hospital admissions, outpatient clinic visits, laboratory results and radiology have been recorded in a computer-based clinical management system since 1996. The hospital records of the index acute MI were reviewed to confirm the authenticity of the diagnosis. Demographic data, CV risk factors, and medications were recorded at baseline. The index date was defined as the date of the first occurrence of acute MI. For registration of outcome during follow-up, a blanking period of 14 days after the index date was applied given that the occurrence of recurrent ACS, ischemic stroke or CV death in the first few days of diagnosis of acute MI was most likely related to the initial presentation rather than a new event.

TRS 2°P

The TRS 2°P is a simple 9-point risk stratification tool to predict CV death, MI, and ischemic stroke.¹⁷ TRS 2°P was calculated at baseline: CHF, 1 point; hypertension, 1 point; age ≥75 years, 1 point; diabetes mellitus, 1 point; prior stroke, 1 point; prior CABG, 1 point; PAD, 1 point; eGFR <60 mL/min/1.73 m², 1 point, and smoking, 1 point, as previously described.¹⁷ Subsequent occurrence of risk factors that could contribute to TRS 2°P were not taken into account. Patients were stratified according to TRS 2°P.

Outcomes, Variables and Data Source

The primary outcome was a composite of (1) non-fatal MI; (2) non-fatal ischemic stroke; and (3) CV death. Secondary outcomes included CV death, non-fatal MI, fatal MI, non-fatal ischemic stroke and fatal ischemic stroke. MI was defined as a clinical scenario consistent with MI requiring hospitalization, confirmed by the presence of either electrocardiograph evidence or cardiac marker evidence. Ischemic stroke was defined as a neurological deficit of sudden onset, persisting >24 h, corresponding to a vascular territory and not explained by other causes (e.g., trauma, infection) with neuroimaging evidence, either computed tomography or magnetic resonance imaging confirming the diagnosis. CV death was defined as death that occurred during hospitalization for a documented MI, stroke, other non-cerebral, non-coronary vascular disease, HF, and/or ventricular tachyarrhythmia, in which there was no conclusive evidence of another cause of death; as well as sudden cardiac death classified according to the Modified Hinkle-Thaler scheme.¹⁹^,²⁰ All patients were monitored until the follow-up date in the cardiac outpatient clinic or until they died. Data were retrieved from the medical records and discharge summaries from the territory-wide information network of all public hospitals in Hong Kong. Patients who were lost to follow-up were contacted by phone. In addition, survival data were obtained from the Births and Deaths General Register Office.

Statistical Analysis

Continuous and discrete variables are expressed as mean±SD and percentages, respectively. Statistical comparison of the baseline clinical characteristics was performed using Student’s t-test or 1-way ANOVA as appropriate. Kaplan-Meier survival analyses with the log-rank test were carried out, and Cox proportional hazards regression modeling was used to calculate hazard ratios (HR) of some predictive factors and their 95% CI for the incidence of different outcomes. The diagnostic performance of the TRS 2°P for the primary composite endpoint and secondary endpoints was further assessed using the c-statistic (area under the curve). The c-statistic for the receiver operating characteristic curve was calculated using Analyze-It for Excel with the Delong-Delong comparison for c-statistic. The c-statistic integrates measures of sensitivity and specificity of the range of a variable. Ideal prediction yields a c-statistic of 1.00, whereas <0.5 indicates that the prediction is no better than chance. Calculations were performed using SPSS (version 21.0) and MedCalc (version 13.1.2). All tests were 2-sided, and P<0.05 was considered significant.

Results

From February 2004 to January 2011, 1,931 patients were hospitalized with a documented MI, of whom 243 died during the index hospitalization. As a result, the final analysis involved 1,688 patients hospitalized for acute MI. Table 1 summarizes their clinical characteristics, inpatient revascularization, and discharge medications. Mean age was 70.3±13.6 years with a male predominance (63.1%). Of note, 57.8% of patients had hypertension, 10.2% had prior CAD requiring revascularization, and 14.4% had had a previous stroke. In addition, 45.3% had eGFR <60 mL/min/1.73 m². A total of 797 patients (47.2%) presented with acute ST-elevation MI (STEMI) and the remaining 891 patients (52.8%) with non-ST-elevation MI (NSTEMI). Mean TRS 2°P was 2.6±1.6. Inpatient coronary revascularization including both percutaneous coronary intervention and CABG was performed in 738 patients (43.7%). Upon discharge, most patients were prescribed aspirin (89.5%), 54.4% were prescribed a P2Y112 inhibitor, and 2.9%, warfarin. In addition, 71.3% of patients were prescribed a statin; 69.9%, a β-adrenergic blocker, and 66.2% an angiotensin-converting enzyme inhibitor (ACEI; Table 1, Supplementary Table 1).

Table 1. Baseline Characteristics, In-Patient Revascularization and Discharge Medication

Demographics	(n=1,688)
Age (years)	70.3±13.6
Male	1,065 (63.1)
Hypertension	976 (57.8)
Diabetes mellitus	600 (35.5)
Smoker	767 (45.4)
STEMI	797 (47.2)
NSTEMI	891 (52.8)
Heart failure	135 (8.0)
PAD	65 (3.9)
Prior coronary revascularization	172 (10.2)
Prior stroke	243 (14.4)
Prior CABG	86 (5.1)
eGFR <60 mL/min/1.73 m²	765 (45.3)
Mean TRS 2°P	2.6±1.6
TRS 2°P
0	126 (7.5)
1	328 (19.4)
2	394 (23.3)
3	366 (21.7)
4	262 (15.5)
5	140 (8.3)
6–9	72 (4.3)
Discharge medication and inpatient revascularization
Aspirin	1,510 (89.5)
P2Y12 inhibitor	918 (54.4)
Warfarin	49 (2.9)
ACEI	1,118 (66.2)
β-blocker	1,180 (69.9)
CCB	274 (16.2)
Statin	1,204 (71.3)
Inpatient PCI	699 (41.4)
Inpatient CABG	39 (2.3)

Data given as mean±SD or n (%). ACEI, angiotensin-converting enzyme inhibitor; CABG, coronary artery bypass grafting; CCB, calcium channel blocker; eGFR, estimated glomerular filtration rate; NSTEMI, non-ST-elevation myocardial infarction; PAD, peripheral artery disease; PCI, percutaneous coronary intervention; STEMI, ST-elevation myocardial infarction; TRS 2°P, TIMI (Thrombolysis in Myocardial Infarction) Risk Score for Secondary Prevention score.

Primary Composite Outcomes

After a mean follow-up of 41.5±34.4 months, 405 patients (24.0%) developed a primary composite endpoint with an annual incidence of 9.3% per year. Table 2 summarizes the factors that were predictive of a primary composite endpoint together with the corresponding HR based on Cox proportional hazard modeling and 95% CI. On univariate analysis, age ≥75 years, female gender, hypertension, diabetes mellitus, PAD, history of HF, prior ischemic stroke, renal impairment with eGFR <60 mL/min/1.73 m², NSTEMI at the index MI and the prescription of calcium channel blocker (CCB) upon discharge were associated with an increasing risk for a primary composite endpoint. In addition, the prescription of aspirin, P2Y12 inhibitor, ACEI, β-adrenergic blocker, statin, and in-hospital revascularization were associated with lower risk of primary composite endpoint. There was a progressive increase in the annual incidence of primary composite endpoint with increasing TRS 2°P. Figure 1A shows a Kaplan-Meier analysis of composite endpoint in patients stratified according to TRS 2°P (log-rank, 222; P<0.0001). For instance, the annual incidence of primary composite endpoint for patients with TRS 2°P 0 was 1.0% per year, increasing to 39.9% per year for those with TRS 2°P ≥6 (HR, 27.6; 95% CI: 9.87–77.39, P<0.001; Figure 1B; Table 2).

Table 2. Baseline Indicators of Composite Endpoint (Cardiovascular Death, Non-Fatal MI, and Non-Fatal Stroke)

	No. composite endpoints	Univariate analysis		Multivariate analysis^†		Multivariate analysis^‡
	No. composite endpoints	HR (95% CI)	P-value	HR (95% CI)	P-value	HR (95% CI)	P-value
Age ≥75 years	290	3.65 (2.94–4.53)	<0.001*	2.00 (1.57–2.55)	<0.001*	1.45 (1.06–1.99)	0.02*
Male	200	0.53 (0.43–0.64)	<0.001*	1.05 (0.86–1.30)	0.634	0.95 (0.76–1.19)	0.647
Hypertension	288	1.96 (1.59–2.45)	<0.001*	1.38 (1.10–1.73)	0.006*	1.01 (0.76–1.35)	0.912
Diabetes mellitus	183	1.62 (1.33–1.97)	<0.001*	1.26 (1.02–1.55)	0.031*	0.99 (0.75–1.30)	0.94
Smoker	174	0.90 (0.74–1.10)	0.304
NSTEMI	281	2.25 (1.82–2.78)	<0.001*	1.16 (0.92–1.46)	0.222	1.16 (0.92–1.47)	0.205
Heart failure	69	3.21 (2.47–4.16)	<0.001*	1.34 (1.02–1.76)	0.036*	1.14 (0.82–1.59)	0.435
PAD	32	2.72 (1.90–3.91)	<0.001*	1.29 (0.89–1.88)	0.181	1.11 (0.73–1.71)	0.621
Prior coronary revascularization	51	1.28 (0.95–1.71)	0.103
Prior stroke	105	2.52 (2.02–3.15)	<0.001*	1.33 (1.05–1.68)	0.018*	1.08 (0.79–1.41)	0.649
Prior CABG	20	0.91 (0.58–1.43)	0.683
eGFR <60 mL/min/1.73 m²	280	3.22 (2.61–3.98)	<0.001*	1.71 (1.35–2.16)	<0.001*	1.23 (0.91–1.67)	0.186
Medications
Aspirin	316	0.25 (0.20–0.32)	<0.001*	0.47 (0.36–0.61)	<0.001*	0.48 (0.37–0.62)	<0.001
P2Y12 inhibitor	125	0.31 (0.25–0.38)	<0.001*	0.71 (0.53–0.94)	0.018*	0.69 (0.52–0.92)	0.011*
Warfarin	11	0.83 (0.45–1.50)	0.53
ACEI	217	0.48 (0.38–0.58)	<0.001*	0.86 (0.70–1.07)	0.178	0.83 (0.67–1.03)	0.084
β-blocker	234	0.50 (0.41–0.61)	<0.001*	0.84 (0.68–1.04)	0.11	0.86 (0.69–1.06)	0.159
Statin	234	0.46 (0.38–0.56)	<0.001*	0.94 (0.75–1.18)	0.600	0.95 (0.75–1.19)	0.629
CCB	81	1.34 (1.05–1.71)	0.018*	0.76 (0.59–0.98)	0.034*	0.75 (0.58–0.96)	0.024*
Inpatient PCI	77	0.28 (0.22–0.35)	<0.001*	0.61 (0.44–0.85)	0.004*	0.62 (0.44–0.87)	0.005*
Inpatient CABG	1	0.09 (0.01–0.62)	0.015*	0.10 (0.01–0.69)	0.02*	0.07 (0.01–0.49)	0.008*
TRS 2°P
0	4	Ref.		Ref.		Ref.
1	29	2.84 (1.00–8.09)	0.050			2.61 (0.91–7.48)	0.075
2	63	5.35 (1.95–14.70)	0.001*			3.99 (1.37–11.64)	0.011*
3	104	10.46 (3.85–28.39)	<0.001*			6.22 (2.00–19.33)	0.002*
4	97	14.60 (5.37–39.70)	<0.001*			7.21 (2.10–24.71)	0.002*
5	69	22.35 (8.16–61.27)	<0.001*			8.46 (2.21–32.30)	0.002*
6–9	39	27.6 (9.87–77.39)	<0.001*			9.46 (2.14–41.82)	0.003*

*P<0.05; ^†parameters with P<0.05 on univariate analysis except TRS 2°P; ^‡parameters with P<0.05 on univariate analysis together with TRS 2°P. MI, myocardial infarction. Other abbreviations as in Table 1.

Figure 1.

Kaplan-Meier estimates of (A) primary composite endpoint-free survival; (C) cardiovascular death-free survival; (E) non-fatal myocardial infarction (MI)-free survival; (G) MI-free survival; (I) non-fatal ischemic stroke-free survival; and annual incidence of (B) primary composite endpoint; (D) cardiovascular death; (F) non-fatal MI; (H) MI; and (J) non-fatal ischemic stroke in post-MI patients stratified according to Thrombolysis In Myocardial Infarction (TIMI) Risk Score for Secondary Prevention (TRP 2°P).

On the first multivariate analysis that did not incorporate TRS 2°P into the model, age ≥75 years, hypertension, diabetes, history of HF, prior ischemic stroke, and renal impairment with eGFR <60 mL/min/1.73 m² at the index MI remained independently associated with increasing risk of primary composite endpoint. In contrast, prescription of aspirin, P2Y12 inhibitors, CCB, and in-hospital revascularization were associated with lower risk of primary composite endpoint. In the second multivariate analysis that included TRS 2°P in the model, age ≥75 years, and TRS 2°P remained independently associated with the increasing risk of primary composite endpoint; and the prescription of aspirin, P2Y12 inhibitor, and/or CCB as well as in-hospital revascularization remained independently associated with lower risk of primary composite endpoint. TRS 2°P was associated with a much higher HR (Table 2). The sensitivity of the TRS 2°P score (76.3% with optimal cut-off at 3) was the highest compared with other baseline demographic factors as a single criterion (sensitivity range, 72.1–80.5%). The c-statistic of TRS 2°P as a predictor of a primary composite endpoint was 0.724 (95% CI: 0.697–0.752; Supplementary Table 2).

In an additional analysis stratifying the risk of primary composite endpoint separately for patients with and without in-hospitalization revascularization, although patients receiving in-hospital revascularization had an overall lower risk of primary composite endpoint compared with those without in-hospital revascularization (3.5%/year vs. 15.3%/year), such risk increased progressively with increasing TRS 2°P in both groups (Supplementary Figure A).

CV Death

There were 278 CV deaths during the study period. The annual CV mortality was 5.86% per year. Similar to the primary composite endpoint, TRS 2°P together with age ≥75 years, history of HF, and renal impairment with eGFR <60 mL/min/1.73 m², and NSTEMI at index MI were associated with increasing risk of CV death both on univariate analysis and on multivariate analysis (Table 3). Figure 1C shows a Kaplan-Meier analysis of CV death for patients stratified according to TRS 2°P (log-rank, 197, P<0.0001). The annual CV mortality for patients with TRS 2°P 0 was 0.2% per year, increasing to 25.8% per year for those with TRS 2°P ≥6 (HR, 83.09; 95% CI: 11.33–609.56, P<0.001; Figure 1D; Table 3). In an additional analysis stratifying the risk of CV death separately for patients with and without in-hospitalization revascularization, while patients receiving in-hospital revascularization had an overall lower risk of CV death compared with those without in-hospital revascularization (1.9%/year vs. 9.7%/year), the risk of CV death increased progressively with increasing TRS 2°P in both groups (Supplementary Figure B). The association between TRS 2°P and CV death remained highly significant after multivariate analysis (Table 3). The sensitivity of TRS 2°P (81.0% with optimal cut-off at 3) was highest compared with other baseline demographic factors as a single criterion to predict CV death (sensitivity, 76.4–85.6%). The c-statistic of TRS 2°P as a predictor of CV death was 0.74 (95% CI: 0.711–0.769; Supplementary Table 3).

Table 3. Baseline Indicators of Cardiovascular Death

	No. cardiovascular deaths	Univariate analysis		Multivariate analysis^†		Multivariate analysis^‡
	No. cardiovascular deaths	HR (95% CI)	P-value	HR (95% CI)	P-value	HR (95% CI)	P-value
Age ≥75 years	218	5.10 (3.84–6.77)	<0.001*	2.70 (1.98–3.68)	<0.001*	1.81 (1.22–2.68)	0.003*
Male	137	0.54 (0.43–0.68)	<0.001*	1.20 (0.94–1.53)	0.154	1.07 (0.82–1.40)	0.695
Hypertension	204	2.16 (1.66–2.82)	<0.001*	1.50 (1.13–1.98)	0.005*	1.03 (0.73–1.46)	0.706
Diabetes mellitus	121	1.45 (1.14–1.83)	0.002*	1.11 (0.87–1.43)	0.405	0.83 (0.61–1.17)	0.288
Smoker	122	0.95 (0.75–1.20)	0.641
NSTEMI	196	2.28 (1.76–2.94)	<0.001*	0.99 (0.75–1.30)	0.917	1.00 (0.76–1.32)	0.016*
Heart failure	54	3.67 (2.73–4.95)	<0.001*	1.42 (1.04–1.94)	0.027*	1.14 (0.78–1.67)	0.047*
PAD	23	2.73 (1.78–4.18)	<0.001*	1.20 (0.77–1.86)	0.423	0.95 (0.58–1.57)	0.512
Prior coronary revascularization	35	1.26 (0.89–1.80)	0.196
Prior stroke	79	2.81 (2.17–3.65)	<0.001*	1.41 (1.07–1.85)	0.014*	1.06 (0.73–1.54)	0.054
Prior CABG	16	1.09 (0.66–1.81)	0.726
eGFR <60 mL/min/1.73 m²	202	3.65 (2.81–4.75)	<0.001*	1.76 (1.32–2.34)	<0.001*	1.19 (0.82–1.71)	0.019*
Medications
Aspirin	200	0.18 (0.14–0.23)	<0.001*	0.36 (0.27–0.48)	<0.001*	0.37 (0.27–0.49)	<0.001*
P2Y12 inhibitor	69	0.24 (0.18–0.31)	<0.001*	0.62 (0.43–0.88)	<0.001*	0.58 (0.41–0.84)	0.003*
Warfarin	9	1.01 (0.52–1.97)	0.97
ACEI	138	0.40 (0.32–0.51)	<0.001*	0.61 (0.47–1.04)	0.097	0.76 (0.59–0.99)	0.038*
β-blocker	138	0.34 (0.27–0.44)	<0.001*	0.61 (0.47–0.79)	<0.001*	0.62 (0.48–0.80)	<0.001*
Statin	142	0.35 (0.28–0.45)	<0.001*	0.88 (0.67–1.16)	0.371	0.89 (0.68–1.17)	0.395
CCB	51	0.34 (0.87–1.57)	0.34
Inpatient PCI	43	0.22 (0.16–0.31)	<0.001*	0.65 (0.42–1.00)	0.048*	0.67 (0.43–1.03)	0.065
Inpatient CABG	1	0.13 (0.19–0.94)	0.043*	0.20 (0.03–1.47)	0.114	0.13 (0.02–0.97)	0.046*
TRS 2°P
0	1	Ref.				Ref.
1	13	4.98 (0.65–38.05)	0.122			4.56 (0.59–35.13)	0.146
2	39	12.95 (1.78–94.27)	0.011*			9.03 (1.18–69.04)	0.034*
3	74	28.60 (3.98–205.70)	0.001*			16.07 (2.01–128.54)	0.009*
4	69	39.27 (5.45–282.77)	<0.001*			17.28 (1.99–149.64)	0.010*
5	53	64.58 (8.93–467.05)	<0.001*			22.64 (2.39–214.18)	0.007*
6–9	30	83.09 (11.33–609.56)	<0.001*			27.82 (2.61–297.21)	0.006*

*P<0.05; ^†parameters with P<0.05 on univariate analysis except TRS 2°P; ^‡parameters with P<0.05 on univariate analysis together with TRS 2°P. Abbreviations as in Table 1.

There were 270 recurrent (fatal and non-fatal) MI with an annual incidence of 6.1%. Of these, 136 MI were fatal, and the remaining 134 were non-fatal. The annual incidence of non-fatal MI was 3.03%. Figure 1E,G shows Kaplan-Meier analyses of non-fatal MI, and fatal and non-fatal MI in patients stratified according to TRS 2°P (log-rank, 53.4, P<0.0001; and log-rank, 151, P<0.0001). Similar to the primary composite endpoint and CV death, increasing TRS 2°P was associated with a progressively higher incidence of MI. For instance, the annual incidence of non-fatal MI in patients with TRS 2°P 0 was 0.8% per year, increasing to 12.9% per year for those with TRS 2°P ≥6 (HR, 10.73; 95% CI: 3.06–37.69, P<0.001; Figure 1F; Table 4). Likewise, the annual incidence of fatal and non-fatal MI for patients with TRS 2°P 0 was 0.8% per year, increasing to 32.8% per year for those with TRS 2°P ≥6. (HR, 28.44; 95% CI: 8.72–92.78, P<0.001; Figure 1H; Supplementary Table 4). In an additional analysis stratifying the risk of non-fatal MI separately for patients with and without in-hospitalization revascularization, while patients receiving in-hospital revascularization had an overall lower risk of non-fatal MI compared with those without in-hospital revascularization (1.4%/year vs. 4.6%/year), the risk of non-fatal MI nonetheless increased progressively with increasing TRS 2°P in both groups (Supplementary Figure C). The sensitivity of TRS 2°P (76.7% with optimal cut-off at 3) was highest compared with other baseline demographic factors as a single criterion to predict non-fatal MI and fatal and non-fatal MI (sensitivity, 71.6–81.7%). The c-statistic of TRS 2°P as a predictor of MI was 0.709 (95% CI: 0.676–0.742; Supplementary Table 5).

Table 4. Baseline Indicators of Non-Fatal MI

	No. non- fatal MI	Univariate analysis		Multivariate analysis^†		Multivariate analysis^‡
	No. non- fatal MI	HR (95% CI)	P-value	HR (95% CI)	P-value	HR (95% CI)	P-value
Age ≥75 years	84	2.34 (1.65–3.32)	<0.001*	1.49 (1.01–2.20)	0.046*	1.44 (0.87–2.40)	0.159
Male	68	0.56 (0.40–0.79)	0.001*	0.92 (0.64–1.32)	0.654	0.92 (0.61–1.36)	0.662
Hypertension	96	1.96 (1.34–2.85)	<0.001*	1.26 (0.85–1.88)	0.258	1.20 (0.72–1.99)	0.485
Diabetes mellitus	73	2.27 (1.62–3.19)	<0.001*	1.74 (1.22–2.50)	0.002*	1.71 (1.06–2.77)	0.028*
Smoker	57	0.88 (0.62–1.24)	0.46
NSTEMI	94	2.28 (1.57–3.30)	<0.001*	1.38 (0.91–2.09)	0.125	1.39 (0.92–2.10)	0.122*
Heart failure	17	2.06 (1.24–3.42)	0.006*	1.05 (0.61–1.78)	0.868	1.07 (0.58–1.98)	0.838
PAD	13	3.18 (1.79–5.63)	<0.001*	1.73 (0.95–3.15)	0.071	1.82 (0.89–3.72)	0.103
Prior coronary revascularization	20	1.55 (0.96–2.49)	0.072
Prior stroke	30	1.97 (1.31–2.96)	0.001*	1.17 (0.76–1.79)	0.481	1.19 (0.68–2.07)	0.546
Prior CABG	5	0.70 (0.29–1.71)	0.43
eGFR <60 mL/min/1.73 m²	87	2.55 (1.79–3.63)	<0.001*	1.46 (0.98–2.18)	0.063	1.38 (0.82–2.33)	0.232
Medications
Aspirin	121	0.68 (0.38–1.21)	0.187
P2Y12 inhibitor	55	0.50 (0.35–0.70)	<0.001*	0.97 (0.62–1.52)	0.896	0.98 (0.62–1.53)	0.919
Warfarin	0	0.05 (0.00–3.38)	0.161
ACEI	80	0.63 (0.45–0.89)	0.009*	0.97 (0.67–1.40)	0.868	0.96 (0.66–1.38)	0.818
β-blocker	101	1.13 (0.76–1.67)	0.557
Statin	95	0.82 (0.56–1.18)	0.281
CCB	36	1.94 (1.33–2.84)	0.001*	1.17 (0.79–1.74)	0.455	1.16 (0.78–1.74)	0.469
Inpatient PCI	32	0.38 (0.26–0.57)	<0.001*	0.63 (0.37–1.06)	0.083	0.63 (0.37–1.07)	0.088
Inpatient CABG	0	0.21 (0.00–5.35)	0.206
TRS 2°P
0	3	Ref.				Ref.
1	14	1.81 (0.52–6.31)	0.349			1.49 (0.42–5.28)	0.538
2	17	1.90 (0.56–6.47)	0.307			1.01 (0.26–3.91)	0.987
3	34	4.34 (1.33–14.12)	0.015*			1.53 (0.35–6.73)	0.572
4	33	6.08 (1.86–19.82)	0.003*			1.51 (0.28–8.07)	0.633
5	20	7.76 (2.31–26.13)	0.001*			1.37 (0.20–9.22)	0.747
6–9	13	10.73 (3.06–37.69)	<0.001*			1.26 (0.14–11.52)	0.837

*P<0.05; ^†parameters with P<0.05 on univariate analysis except TRS 2°P; ^‡parameters with P<0.05 on univariate analysis together with TRS 2°P. Abbreviations as in Tables 1,2.

Ischemic Stroke

During the study period there were 49 ischemic strokes, of which 33 were non-fatal and 16 were fatal. Figure 1I shows Kaplan-Meier analyses of non-fatal ischemic stroke for patients stratified according to TRS 2°P (log-rank, 3.19, P=0.074). The annual incidence of non-fatal stroke was 0.71% per year. On univariate analysis, only female gender and renal impairment with eGFR <60 mL/min/1.73 m² were associated with increasing risk of non-fatal ischemic stroke (Supplementary Table 6). Although the annual incidence of non-fatal ischemic stroke appeared to increase with TRS 2°P, the association did not reach statistical significance (Figure 1J; Supplementary Table 6). Furthermore, patients receiving in-hospital revascularization had an overall lower risk of ischemic stroke compared with those without in-hospital revascularization (0.3%/year vs. 1.1%/year; Supplementary Figure D). In addition, there were altogether 18 hemorrhagic strokes during the study period. The annual incidence was 0.38% per year. Nonetheless, there was no statistically significant difference across TRS 2°P strata.

Discussion

To the best of our knowledge, this is the first study to validate TRS 2°P for stratifying patients who survive MI for risk of recurrent CV events in a real-world clinical setting. We first showed that patients who survive MI were at high risk of subsequent major CV events, with an annual incidence of a primary composite endpoint (CV death, non-fatal MI and non-fatal stroke) of 9.3%. This high annual incidence was primarily driven by the high annual incidence of CV death (5.86%) and non-fatal MI (3.03%), rather than non-fatal stroke (0.71%). Second, we demonstrated a strong gradient of primary composite endpoint stratified using TRS 2°P in a real-world cohort of patients who survived MI and in concordance with previous analyses using data from 2 large RCT. The annual incidence of primary composite endpoints in patients with TRS 2°P 0 was 1.0% per year, and progressively increased to 39.9% per year in those with TRS 2°P ≥6, nearly a 40-fold gradient increase. The diagnostic performance of TRS 2°P to predict a primary composite endpoint appears to be reasonable for clinical application (c-statistic, 0.724; 95% CI: 0.697–0.752). Third, TRS 2°P demonstrated similar gradients for the risk of CV death and non-fatal MI to that for the primary composite endpoint, but not for non-fatal stroke, presumably due to the low annual incidence.

Current international guidelines emphasize the need for total CV risk assessment in order to assign preventive measures of appropriate intensity according to an individual patient’s risk.¹^,²¹ The recommended risk assessment schemes typically define CV risk on the basis of risk score in an apparently healthy person (primary prevention), or on the basis of existing CV disease (secondary prevention). Unlike the primary prevention setting in which patients are stratified into different risk categories for preventive measures of different intensities, those who survive MI are homogeneously classified into a single “very high-risk” category, and warrant the most aggressive secondary preventive measures.¹^,²¹ In the past decade, treatment options and strategies for secondary prevention in patients who survive MI have expanded rapidly. Although effective, these novel therapeutic options and strategies are expensive and might be associated with some adverse effects. Further sub-classification of these very high-risk patients according to individual absolute risk for major CV events is therefore necessary to enable efficient and appropriate use of these new agents.

Until recently, there has been no validated risk assessment scheme to guide secondary prevention. Using data from the TIMI 50 trial¹⁶ and IMPROVE-IT trial,¹⁷ TRS 2°P has been developed and validated to predict recurrent CV events in patients with previous coronary events. Indeed, TRS 2°P identifies an approximately 5-fold gradient in the risk of subsequent major CV events in post-coronary event patients across low, intermediate-, and high-risk categories.¹⁶^,¹⁷ Similar to the CHA₂ DS₂-VASc score for stroke risk stratification in the setting of atrial fibrillation,²²^–²⁷ TRS 2°P consists of readily available clinical characteristics. This allows easy clinical translation and rapid, widespread application in real-world clinical practice after further validation using real-world clinical data. Indeed, in agreement with previous analyses using RCT data,¹⁶^,¹⁷ the TRS 2°P score likewise identifies a strong gradient of risk for subsequent CV events in a real-world, unselected cohort of patients who survive MI. Nonetheless quantitative differences exist between the present real-world cohort and that of the 2 previous RCT. Compared with corresponding annualized rates of major CV events from previous analyses,¹⁶^,¹⁷ patients with low TRS 2°P (0–1) in the present study were at similar risk of major CV events, but those with an intermediate score (TRS 2°P 2) or high score (TRS 2°P ≥3) had a much higher risk of major CV events (2–3-fold; Figure 2).

Figure 2.

Annual risk of primary composite endpoint stratified according to Thrombolysis In Myocardial Infarction (TIMI) Risk Score for Secondary Prevention (TRP 2°P) from the 2 randomized control trials: IMPROVE-IT¹⁷ and TIMI-50,¹⁶ and the real-world registry from the present study. CABG, coronary artery bypass grafting; CHF, congestive heart failure; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; HT, hypertension; PAD, peripheral artery disease.

Obviously, patients from these RCT¹⁶^,¹⁷ were highly selected based on stringent inclusion and exclusion criteria, and were often not comparable with the more heterogeneous patients in the real-world registries. In particular there were fewer patients at the extremely high-risk end of the spectrum. Of note, patients from the 2 studies: IMPROVE IT trial and TIMI 50 trial, were substantially younger, 63.6 years¹⁷ and 59 years,¹⁶ respectively, than those in the present study, 70.3 years. Equally important, much higher proportions of patients in the 2 studies were prescribed guideline-recommended medical therapy such as P2Y12 inhibitor, ACEI/angiotensin receptor blocker, β-adrenergic blocker or statin¹⁶ compared with the current study. Furthermore, the differences in type of index coronary event and the duration from the index acute coronary event to study recruitment may at least partly explain the much higher event rate in the present study. In the IMPROVE-IT study, although all patients had had recent ACS (within 10 days), only 29% had STEMI, and 47% had an NSTEMI, with the remaining 24% having only unstable angina.¹⁷ In the TIMI-50 trial, although all patients had a history of spontaneous MI, only 45% occurred in the 3 months prior to recruitment.¹⁶ In the present real-world registry, all patients who survived MI (both STEMI and NSTEMI) and who were discharged from hospital during the study period were included in the final analysis, thereby providing an inclusive cohort to estimate the natural risk of MI. Nonetheless, TRS 2°P in both a clinical trial setting and in the present real-world registry has consistently demonstrated a high discriminating power to identify extremely high-risk post-coronary event patients (TRS 2°P >3).

Expectedly, the ability to further stratify the CV risk of individual patients who survive an MI enables more quantitative decision-making, which involves a delicate balance between potential benefits and adverse effects, and sometimes cost, thereby facilitating treatment allocation. For instance, using the TIMI 50 trial data,¹⁶ a gradient of absolute risk reduction with vorapaxar across TRS 2°P was clearly demonstrated; patients at a higher absolute risk could derive greater absolute benefit than those at lower risk from the same effective treatment. The much greater absolute risk reduction in patients with high TRS 2°P outweighs the accompanying bleeding from vorapaxar, thereby justifying the use of such treatment.¹⁶ Likewise, in the analysis of the IMPROVE-IT trial data,¹⁷ patients with TRS 2°P >3 derived a 6.6% absolute reduction in risk of CV death, major coronary event, or stroke from the addition of ezetimibe to simvastatin therapy, whereas patients with the lowest risk (TRS 2°P 0–1) had no such reduction from ezetimibe over the study period of 7 years. Furthermore, in the present study, patients receiving in-hospital revascularization had an overall lower risk of primary composite endpoint, CV death, non-fatal MI and ischemic stroke compared with patients who did not undergo in-hospital revascularization. Patients with higher TRS 2°P had a higher absolute risk reduction in the primary composite endpoint compared with patients with lower TRS 2°P. It has been shown in the Timing of Intervention in ACS (TIMACS) trial involving patients with ACS that early intervention compared with late intervention reduced the rate of the composite outcome of death, MI, or refractory ischemia, particularly in high-risk patients.²⁸ Nonetheless, it is important to stress that the timing of coronary intervention in the present study was not chosen in a randomized fashion, and it is possible that residual confounding may be evident such that patients with and without in-hospital revascularization were in some ways different from each other.

Study Limitations

This study was limited by its registry-based and single-center observational design in primarily hospital-based patients. Due to the observational nature, the use of guideline-recommended medical therapy was suboptimal compared with RCT; nonetheless it represents a typical pattern of real-world clinical practice. In addition, data pertaining to medication adherence, another important factor to determine outcome, were unfortunately not available in the present study. Although we carefully ascertained all MI and strokes by careful examination of hospital records, laboratory and imaging results, patients with a milder form of MI and/or stroke who were not hospitalized were not included.

Conclusions

Taken collectively, the present study provides additional supporting evidence using real-world clinical data for the use of TRS 2°P to stratify patients with established CV disease for future risk of CV events.

Disclosures

The authors declare no conflicts of interest.

Supplementary Files

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-18-0308

References

1. Stone NJ, Robinson JG, Lichtenstein AH, Bairey Merz CN, Blum CB, Eckel RH, et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2014; 63: 2889–2934.
2. Ibanez B, James S, Agewall S, Antunes MJ, Bucciarelli-Ducci C, Bueno H, et al. 2017 ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation: The Task Force for the management of acute myocardial infarction in patients presenting with ST-segment elevation of the European Society of Cardiology (ESC). Eur Heart J 2018; 39: 119–177.
3. Drozda JP Jr, Ferguson TB Jr, Jneid H, Krumholz HM, Nallamothu BK, Olin JW, et al. 2015 ACC/AHA Focused Update of Secondary Prevention Lipid Performance Measures: A Report of the American College of Cardiology/American Heart Association Task Force on Performance Measures. J Am Coll Cardiol 2016; 67: 558–587.
4. Bonaca MP, Bhatt DL, Cohen M, Steg PG, Storey RF, Jensen EC, et al. Long-term use of ticagrelor in patients with prior myocardial infarction. N Engl J Med 2015; 372: 1791–1800.
5. Cannon CP, Blazing MA, Giugliano RP, McCagg A, White JA, Theroux P, et al. Ezetimibe added to statin therapy after acute coronary syndromes. N Engl J Med 2015; 372: 2387–2397.
6. Lloyd-Jones DM, Morris PB, Ballantyne CM, Birtcher KK, Daly DD Jr, DePalma SM, et al. 2017 Focused Update of the 2016 ACC Expert Consensus Decision Pathway on the Role of Non-Statin Therapies for LDL-Cholesterol Lowering in the Management of Atherosclerotic Cardiovascular Disease Risk: A Report of the American College of Cardiology Task Force on Expert Consensus Decision Pathways. J Am Coll Cardiol 2017; 70: 1785–1822.
7. Sabatine MS, Giugliano RP, Keech AC, Honarpour N, Wiviott SD, Murphy SA, et al. Evolocumab and clinical outcomes in patients with cardiovascular disease. N Engl J Med 2017; 376: 1713–1722.
8. Marso SP, Daniels GH, Brown-Frandsen K, Kristensen P, Mann JF, Nauck MA, et al. Liraglutide and cardiovascular outcomes in type 2 diabetes. N Engl J Med 2016; 375: 311–322.
9. Fonarow GC, Keech AC, Pedersen TR, Giugliano RP, Sever PS, Lindgren P, et al. Cost-effectiveness of evolocumab therapy for reducing cardiovascular events in patients with atherosclerotic cardiovascular disease. JAMA Cardiol 2017; 2: 1069–1078.
10. Villa G, Lothgren M, Kutikova L, Lindgren P, Gandra SR, Fonarow GC, et al. Cost-effectiveness of evolocumab in patients with high cardiovascular risk in Spain. Clin Ther 2017; 39: 771–786.e773.
11. Gandra SR, Villa G, Fonarow GC, Lothgren M, Lindgren P, Somaratne R, et al. Cost-effectiveness of LDL-C lowering with evolocumab in patients with high cardiovascular risk in the United States. Clin Cardiol 2016; 39: 313–320.
12. Wilson PW, D’Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation 1998; 97: 1837–1847.
13. Conroy RM, Pyorala K, Fitzgerald AP, Sans S, Menotti A, De Backer G, et al. Estimation of ten-year risk of fatal cardiovascular disease in Europe: The SCORE project. Eur Heart J 2003; 24: 987–1003.
14. Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, May M, Brindle P. Derivation and validation of QRISK, a new cardiovascular disease risk score for the United Kingdom: Prospective open cohort study. BMJ 2007; 335: 136.
15. Morrow DA, Braunwald E, Bonaca MP, Ameriso SF, Dalby AJ, Fish MP, et al. Vorapaxar in the secondary prevention of atherothrombotic events. N Engl J Med 2012; 366: 1404–1413.
16. Bohula EA, Bonaca MP, Braunwald E, Aylward PE, Corbalan R, De Ferrari GM, et al. Atherothrombotic risk stratification and the efficacy and safety of vorapaxar in patients with stable ischemic heart disease and previous myocardial infarction. Circulation 2016; 134: 304–313.
17. Bohula EA, Morrow DA, Giugliano RP, Blazing MA, He P, Park JG, et al. Atherothrombotic risk stratification and ezetimibe for secondary prevention. J Am Coll Cardiol 2017; 69: 911–921.
18. Wong KL, Wong YT, Yung SY, Tam CC, Lam CC, Hai SH, et al. A single centre retrospective cohort study to evaluate the association between implementation of an acute myocardial infarction clinical pathway and clinical outcomes. Int J Cardiol 2015; 182: 82–84.
19. Greenberg H, Case RB, Moss AJ, Brown MW, Carroll ER, Andrews ML. Analysis of mortality events in the Multicenter Automatic Defibrillator Implantation Trial (MADIT-II). J Am Coll Cardiol 2004; 43: 1459–1465.
20. Hinkle LE Jr, Thaler HT. Clinical classification of cardiac deaths. Circulation 1982; 65: 457–464.
21. Catapano AL, Graham I, De Backer G, Wiklund O, Chapman MJ, Drexel H, et al. 2016 ESC/EAS guidelines for the management of dyslipidaemias. Eur Heart J 2016; 37: 2999–3058.
22. Chiang CE, Okumura K, Zhang S, Chao TF, Siu CW, Wei Lim T, et al. 2017 consensus of the Asia Pacific Heart Rhythm Society on stroke prevention in atrial fibrillation. J Arrhythm 2017; 33: 345–367.
23. Kirchhof P, Benussi S, Kotecha D, Ahlsson A, Atar D, Casadei B, et al. 2016 ESC guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Eur Heart J 2016; 37: 2893–2962.
24. Chan PH, Lau CP, Tse HF, Chiang CE, Siu CW. CHA2DS2-VASc recalibration with an additional age category (50–64 years) enhances stroke risk stratification in Chinese patients with atrial fibrillation. Can J Cardiol 2016; 32: 1381–1387.
25. Chan PH, Huang D, Lau CP, Chan EW, Wong IC, Lip GY, et al. Net clinical benefit of dabigatran over warfarin in patients with atrial fibrillation stratified by CHA2DS2-VASc and time in therapeutic range. Can J Cardiol 2016; 32: 1247e1215–1247.e1221.
26. Siu CW, Tse HF. Net clinical benefit of warfarin therapy in elderly Chinese patients with atrial fibrillation. Circ Arrhythm Electrophysiol 2014; 7: 300–306.
27. Siu CW, Lip GY, Lam KF, Tse HF. Risk of stroke and intracranial hemorrhage in 9727 Chinese with atrial fibrillation in Hong Kong. Heart Rhythm 2014; 11: 1401–1408.
28. Mehta SR, Granger CB, Boden WE, Steg PG, Bassand JP, Faxon DP, et al. Early versus delayed invasive intervention in acute coronary syndromes. N Engl J Med 2009; 360: 2165–2175.

Corresponding author

Register with J-STAGE for free!